INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ADVANCES IN GENETICALLY ENGINEERED MICE with overexpression or selective ablation of a particular gene of interest have progressed at an amazing pace over the past few years. Transgenic mouse models have provided unique insights into a wide variety of physiological processes including the field of cardiac electrophysiology. By generating animals in which modified genes or the cognate of cDNAs are placed into the genome and the encoded proteins are then expressed in an organ-specific manner, we have available for the first time the unique opportunity to study the in vivo structure-function relation of a particular ion channel gene of interest. Therefore, information about the electrophysiology of mouse cardiac myocytes is urgently needed to accurately interpret the physiological data from these transgenic mouse models.

Using transgenic mouse models as a tool, important information has been obtained with regard to the molecular correlates of different K⁺ currents in adult mouse myocytes (4, 35, 38); however, there is no information on Ca²⁺-activated Cl currents (I_Cl,Ca). As an important framework for future comparative studies, we have recently examined the basic electrophysiological and pharmacological characteristics of this current in adult mouse ventricular myocytes using whole cell and single channel patch-clamp techniques.

In many types of cardiac cells, a transient outward current (I_to) contributes to the initial phase of repolarization during the action potential. Kenyon and Gibbons (24) have provided evidence that the I_to in sheep Purkinje fibers consists of a voltage-activated 4-aminopyridine (4-AP)-sensitive K⁺ current as well as a smaller Cl-selective current. Subsequent studies (19, 41, 48) revealed that in many cardiac cells, the I_to is composed of both Ca²⁺-insensitive and Ca²⁺-sensitive components. Recent molecular studies (10, 22, 52) have reported the cloning of K⁺ channels from several species including humans, which represent the molecular correlates for the Ca²⁺-insensitive 4-AP-sensitive I_to. In contrast, the identification of the Ca²⁺-sensitive component of the I_to has remained elusive. The current is present in rabbit ventricular (56) and atrial (57) myocytes and Purkinje cells (42). In canine ventricular myocytes, a Ca²⁺-sensitive I_to has been shown to be blocked by anion transport inhibitors, which suggests that the current is carried by Cl ions (55). Single channel studies (6) further confirmed that this current is Cl selective and has a very small conductance (in the range of 1.0 pS).

The potential importance of I_Cl,Ca for cardiac repolarization and as a charge carrier for the arrhythmogenic transient inward current (I_ti) has been suggested (17, 58). However, information on this potentially important current is lacking in mice. Here, we report for the first time the presence of Ca²⁺-activated I_to in mouse ventricular myocytes. Similar to canine cardiac myocytes, the current shows a very small single channel conductance (~1 pS). The channel is critically dependent on Ca²⁺ entry via voltage-gated Ca²⁺ channels and release from intracellular Ca²⁺ stores. However, unlike the current previously reported in canine and rabbit myocytes, we have demonstrated for the first time that the channel also shows weak voltage dependence. Furthermore, we have identified the functional roles of the current in the action potential waveform using perforated-patch techniques; the current may play an important role in the early repolarization of the action potentials.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Preparation of cardiac myocytes. Single mouse ventricular myocytes were isolated as previously described with slight modification (2). Mice were injected with ~0.1 ml heparin (1,000 U/ml) ~10 min before heart excision. Animals were anesthetized with 0.1 ml of pentobarbital sodium (50 mg/ml ip). Hearts were removed, placed into ice-cold nominally Ca²⁺-free buffered solution, cannulated under a dissecting microscope, and mounted on a Langendorff apparatus. Hearts were perfused with Ca²⁺-free modified Tyrode solution composed of (in mM) 140 NaCl, 5.4 KCl, 1 MgCl₂, 10 HEPES, and 10 glucose (pH 7.4 with NaOH). The perfusate was gassed with 100% O₂ and maintained at 37°C. The perfusion pressure was monitored and the flow rate was adjusted to maintain perfusion pressure at ~60 mmHg. After 10 min of perfusion, the perfusion solution was switched to one containing 0.025 mM Ca²⁺, 0.37 mg/ml of collagenase B (Boehringer Mannheim; Mannheim, Germany), and 0.027 mg/ml of protease (Sigma Chemical; St. Louis, MO). After 10-15 min of collagenase perfusion, hearts were removed from the perfusion apparatus and left ventricular tissue was separated from the atria, great vessels, and right ventricle. The left ventricle was minced and incubated in a shaking bath for another 5-10 min in collagenase-containing solution. Cells were then harvested, washed twice, and stored at room temperature using a high-K⁺ solution containing (in mM) 120 K-glutamate, 25 KCl, 1 MgCl₂, 0.1 EGTA, 10 glucose, and 10 HEPES; pH 7.4 with KOH. Cells were used for electrophysiological recording within 7-8 h after isolation. This isolation procedure yields ~80% of Ca²⁺-tolerant ventricular myocytes with clear striations.

Whole cell current recordings. Action potentials were recorded at room temperature using the perforated-patch technique to avoid dialysis of the intracellular milieu (26). All other experiments were performed using the conventional whole patch-clamp technique (16) at room temperature. Patch electrodes were pulled from borosilicate glass and had 2- to 5-M tip resistances. Recordings were done using an Axopatch 200B patch-clamp amplifier (Axon Instruments; Foster City, CA) interfaced to a personal computer. Voltage or current commands and data collection were performed using custom-written software. During voltage-clamp experiments, the cell capacitance was calculated by integrating the area under an uncompensated capacitive transient elicited by a 20-mV hyperpolarizing pulse from a holding potential of 40 mV. Cell capacitance and series resistance were then compensated as much as possible almost to the point of ringing. In general, 60-80% of the series resistance was compensated. Whole cell current records were filtered at 2 kHz and sampled at 10 kHz. Data were stored in the computer for analysis using custom-written software.

Solutions. For action potential recordings, the patch pipettes were backfilled with amphotericin (200 µg/ml) prepared from a fresh stock of 50 mg/ml amphotericin in DMSO. Pipette solution contained (in mM) 120 K-glutamate, 25 KCl, 1 MgCl₂, 1 CaCl₂, and 10 HEPES; pH 7.4 with KOH (Table 1). The inclusion of Ca²⁺ in the solution assured rapid cell death in the event of an inadvertent rupture into a whole cell configuration. The external solution contained (in mM) 138 NaCl, 4 KCl, 1 MgCl₂, 2 CaCl₂, 0.33 NaH₂PO₄, 10 glucose, and 10 HEPES; pH 7.4 with NaOH.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Existence of a 4-aminopyridine (4-AP)-resistant transient outward current (Ito). Examples of a family of Ito recorded from a holding potential of 80 mV using voltage steps from 40 to 100 mV at 10-mV increments at a stimulation frequency of 0.1 Hz in control and after applications of 4-AP (5 mM) and 4-AP plus niflumic acid (Nif, 100 µM). Protocol used is identified above the current traces. Application of 4-AP led to a reduction of a significant component of the Ito. However, there existed a second small component that was blocked by niflumic acid. Current traces shown were normalized to the cell capacitance.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Charge carrier for the 4-AP-resistant Ito. A: examples of current traces recorded using a two-pulse protocol from a holding potential of 55 to 0 mV for 10 ms and a second voltage step to +50 mV in high-Cl vs. low-Cl solutions. Experiments were performed using pipette solution containing 150 mM N-methyl-D-glucamine (NMG), 4 mM Mg-ATP, 1 mM MgCl2, 50 µM BAPTA, and 10 mM HEPES; pH 7.4 with HCl (40 mM) and glutamic acid. External solution contained (in mM) 100 NMG, 20 NaCl/Na-glutamate, 20 tetraethyl ammonium (TEA)-Cl/TEA-OH, 5 4-AP, 2 CaCl2, 1 MgCl2, 10 glucose, and 10 HEPES; pH 7.4 with 1) HCl for high-Cl solution with a calculated Cl reversal potential (ECl) of 33 mV, or 2) glutamic acid for low-Cl solution with a calculated ECl of 47.4 mV. First voltage step elicited a large inward Ca2+ current followed by a second voltage step to +50 mV to decrease the driving force for Ca2+ current. A large 4-AP-resistant Ito could be appreciated in high-Cl solution. In contrast, the Ito was completely abolished in low-Cl solution with ECl of 47.4 mV, which suggests that the charge carrier for the 4-AP-resistant Ito was Cl. B: reversal potentials were determined by tail currents, which were obtained by using a double-pulse protocol consisting of a 50-ms prepulse to +80 mV from a holding potential of 55 mV, followed by 300-ms voltage steps from 40 to +80 mV in 10-mV increments at a stimulation frequency of 0.1 Hz. Tail-current traces were elicited in three different extracellular Cl concentrations ([Cl]o): 128, 82, and 12 mM. Effective series resistances were 0.56, 0.48, and 0.89 M for the examples shown. C: ECl was plotted against different [Cl]o values. Solid line represents the linear regression through the data points with an estimated slope of 59 mV. Data represent means ± SE from 6 cells.

Data analysis. Currents plotted in the current-voltage relations were determined from non-leak-subtracted records. Thus data with substantial leakage current were discarded. Curve fits and data analysis were performed using Origin software (MicroCal; Northampton, MA). Where appropriate, pooled data are presented as means ± SE. Statistical comparison was performed using Student's t-test with P < 0.05 considered significant.

Single channel current recordings. For single channel current recordings, cell-attached patches or excised inside-out patches were used. For cell-attached patches, the bath solution contained (in mM) 120 K-glutamate, 25 KCl, 1 MgCl₂, 0.1 EGTA, 10 glucose, and 10 HEPES; pH 7.4 with KOH to depolarize the membrane potential to 0 mV. The pipette solution contained (in mM) 115 NMG, 5 KCl, 1 MgCl₂, 2 CaCl₂, 20 TEA, 5 4-AP, 10 glucose, and 10 HEPES; pH 7.4 with HCl. For excised patches, we used symmetrical Cl solutions. The bath solution contained (in mM) 140 NMG, 5 CsCl, 2.3 MgCl₂, 1 EGTA, 10 HEPES, and 10 glucose; pH 7.4 with HCl. Pipette solution contained similar composition with the exception that 20 mM NMG was replaced with 5 mM 4-AP and 20 mM TEA. External solution also contained CaCl₂ with calculated pCa of 2, 4, or 5 for different experiments using Calcium Titration software (40). Currents were filtered at 500 Hz and digitized at a frequency of 5 kHz. When filled with the pipette solution, the electrode resistance ranged from 8 to 10 M. To reduce the capacity transient, Sylgard silicone elastomer (Dow Corning; Midland, MI) was applied as close to the pipette tip as possible. Only patches with seal resistances >20-100 G were used. For excised inside-out patches, we used quartz electrodes to obtain a very low noise recording condition. Quartz electrodes were pulled with a laser puller (model P2000, Sutter Instruments; Novato, CA). Leakage and capacity currents were subtracted from unitary current records by fitting a smooth template to null traces. Amplitude histograms at a given test potential were generated and fitted to a single Gaussian distribution using a Levenberg-Marquardt algorithm to obtain the mean unitary currents. Leak-subtracted current records were idealized with a half-height criterion (7). Idealized records were used to construct ensemble-averaged currents and open probability.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Existence of a 4-AP-resistant I_to in mouse ventricular myocytes. Figure 1 shows I_to recorded from mouse ventricular myocytes in control, after 4-AP (5 mM), and after 4-AP plus niflumic acid (100 µM) administration. The intracellular Ca²⁺ buffer (50 µM BAPTA) was kept at a relatively low level in these experiments. A large component of the I_to was blocked by 4-AP. However, there remained a second component of the I_to that was resistant to 4-AP but was blocked by niflumic acid.

Charge carrier of 4-AP-resistant I_to. Close examination of Fig. 1 shows some remaining outward currents after the block by 4-AP and niflumic acid. The remaining outward current after exposure to niflumic acid may represent Ca²⁺-activated nonspecific currents. We performed additional experiments by replacing KCl in the internal solution with NMG as shown in Figs. 2A, 3, and 4. When KCl in the internal solution was replaced by NMG, the I_to in the presence of 4-AP could be completely abolished by low-Cl solution, anion transport blockers, dihydropyridine antagonists, ryanodine, or caffeine.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Pharmacology of the 4-AP-resistant Ito. A and C: examples of whole cell currents recorded using the same two-pulse protocol (inset) and external high-Cl solution and internal solutions as in Fig. 2 with a calculated ECl of 33 mV. First voltage step elicited a large inward Ca2+ current and subsequent second voltage step to +50 mV. A large 4-AP-resistant Ito could be appreciated in control. In contrast, the Ito was completely abolished after application of niflumic acid (50 µM; A) or 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS, 100 µM; C). B and D: summary data showing the transient current (peak current minus sustained current at end of 500-ms voltage steps) elicited at a test potential of +50 mV in control and after niflumic acid (B) or DIDS (D) administration; n = 3 for each group.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Activation of the 4-AP-resistant Ito is dependent on Ca2+ entry via voltage-gated Ca2+ channels and sarcoplasmic reticulum (SR) Ca2+ release. A: examples of whole cell currents elicited in control and after abolishment of the inward Ca2+ currents using Cd2+ (0.5 mM) or nimodipine (5 µM) are shown. Experiments were performed using the same two-pulse protocol (inset) and external and internal solutions as in Fig. 3 with a calculated ECl of 33 mV. First voltage step elicited a large inward Ca2+ current and a second voltage step to +50 mV. Ito was completely abolished after application of Cd2+ or nimodipine. B: summary data showing the transient current elicited at a test potential of +50 mV in control and after Cd2+ or nimodipine administration; n = 3 for each group. C and D: examples of whole cell currents elicited in control and after abolishment of the SR Ca2+ using ryanodine (10 µM) or caffeine (5 mM). Experiments were performed using the same two-pulse protocol (inset) and external and internal solutions as in A. Ito was completely abolished after application of ryanodine or caffeine.

4-AP-insensitive I_to can be blocked by anion transport blockers. We next examined the pharmacology of the identified 4-AP-resistant transient outward Cl currents using different anion transport blockers: niflumic acid and DIDS. One previous study (50) has shown the unexpected blocking effects of the fenamates (e.g., niflumic acid) and the disulfonic stilbenes (DIDS and SITS) on Kv4.3 channels and to a lesser extent Kv4.2 channels, which are the two components that underlie the 4-AP-sensitive I_to. Therefore, we tested the effects of these drugs by first blocking the 4-AP-sensitive I_to with 5 mM 4-AP. Data are summarized in Fig. 3. Current traces in Fig. 3, A and C, were recorded using the same two-pulse protocol in the presence of 5 mM 4-AP before and after application of niflumic acid or DIDS. Figure 3, B and D, shows data that summarize the degree of blockade of the 4-AP-resistant I_to (peak current minus the sustained current at the end of the 500-ms pulse) elicited at a test potential of +50 mV by niflumic acid or DIDS.

Cl current is dependent on Ca²⁺ entry via voltage-gated Ca²⁺ channels and sarcoplasmic reticulum Ca²⁺ stores. We determined the effects of two different Ca²⁺ channel blockers, Cd²⁺ and the dihydropyridine nimodipine, which is a known blocker of the L-type Ca²⁺ channel, on the 4-AP-insensitive I_to. Figure 4A shows current traces elicited using the same two-pulse protocol as in Fig. 2 in control and after Cd²⁺ or nimodipine application. The 4-AP-resistant I_to was completely blocked in the presence of these Ca²⁺ channel blockers. The summary data are shown in Fig. 4B. The data confirm the dependency of the current on Ca²⁺ entry via voltage-gated Ca²⁺ channels. The I_to was also attenuated after application of ryanodine or caffeine to abolish the sarcoplasmic reticulum (SR) Ca²⁺ store. Figure 4, C and D, shows current traces recorded in the presence of 5 mM 4-AP in control and 10 min after application of 10 µM ryanodine or 5 mM caffeine. It is evident that the Ca²⁺-activated I_to is also dependent on the SR Ca²⁺ release. A large 4-AP-resistant I_to could be appreciated in control. In contrast, the I_to was completely abolished after application of ryanodine or caffeine.

Direct demonstration for presence of I_Cl,Ca in mouse ventricular myocytes using single channel recordings. Because anion transport blockers may affect other ionic currents, we used single channel currents to directly demonstrate the presence of I_Cl,Ca. Both cell-attached and excised inside-out configurations were used. Figure 5A shows examples of single channel activities that were recorded under symmetrical Cl conditions upon excision of a patch into the bath solution containing 1 mM Ca²⁺ (pCa = 3; step potentials used are indicated at left). To further confirm the transient nature of the current as recorded from the whole cell conditions, a cell-attached configuration was used (Fig. 5B). A high-K⁺ bath solution was used to depolarize the resting membrane potential to 0 mV. The pipette solution contained 2 mM Ca²⁺, and the holding potential was 55 mV. Single channel outward currents were recorded upon depolarization to various potentials (Fig. 5B). The channels opened with a brief first latency and in general only opened at the beginning of the pulse. Ensemble-averaged currents recorded under this condition faithfully reproduced the whole current kinetics (Fig. 5C). In addition, ensemble-averaged current allowed us to directly calculate the time to peak current without contamination from inward Ca²⁺ current. The time to peak current at +40 mV was estimated to be 70 ± 3.4 ms (n = 3). The single channel current-voltage relation recorded using the cell-attached configuration is shown in Fig. 5D. The solid line represents the least-square fit to the data point yielding a single channel conductance () of 1.2 pS.

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Direct demonstration for the presence of Ca2+-activated Cl currents (ICl,Ca) in mouse ventricular myocytes using single channel recordings. A: examples of current traces recorded using excised inside-out patches in symmetrical Cl condition. Bath solution contained (in mM) 140 NMG, 5 CsCl, 2.3 MgCl2, 1 EGTA, 10 HEPES, and 10 glucose; pH 7.4 with HCl. Pipette solution contained similar composition with the exception that 20 mM NMG was replaced with 5 mM 4-AP and 20 mM TEA. In addition, the bath solution contained Ca2+ with a pCa of 3. Patches were held at 55 mV. Test potentials used are indicated (left of current traces). Single channel activities can be seen on excision of the patch into external solution containing Ca2+. B: examples of single channel currents obtained in cell-attached configuration from a holding potential of 55 mV. High-K+ bath solution was used to depolarize membrane potential to 0 mV. Pipette solution contained (in mM) 115 NMG, 5 KCl, 1 MgCl2, 2 CaCl2, 20 TEA, 5 4-AP, 10 glucose, and 10 HEPES; pH 7.4 with HCl and 2 mM Ca2+. C: ensemble-averaged current at a test potential of +40 mV from a cell-attached patch showing the transient nature of the current. D: corresponding current-voltage relation. Data represent means ± SE from 3 patches with calculated single channel conductance of 1.2 pS.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   ICl,Ca in mouse ventricular myocytes shows weak voltage dependence. A plot of open probability (nPo) against voltage is shown. Solid line represents a single Boltzmann fit to the data points indicating the membrane potential at which half-activation (V1/2) of 46.7 mV and a slope factor of 7.8 occur. Data represent means ± SE (n = 3).

Direct demonstration of the single channel Cl current inhibition by anion transport blockers. To crosscheck our findings that the niflumic acid-sensitive whole cell current indeed represents a second component of the I_to, we examined the effects of the blockers on the single channel currents (Fig. 7). Figure 7A shows current traces recorded at a test potential of +60 mV using cell-attached configuration in control and after niflumic acid (100 µM) administration. There was a rapid disappearance of the single channel activities (n = 3). Figure 7B shows similar findings using an excised patch. In these experiments, the pipette solution contained no Ca²⁺. Appearance of channel activities can be seen only after excision of the patch into the external solution containing Ca²⁺ with a pCa of 3. The channels were rapidly blocked upon bath application of DIDS (100 µM). Figure 7B, bottom, shows open probability in the control and after application of DIDS.

View larger version (25K): [in this window] [in a new window]   Fig. 7.   Direct demonstration of the single channel Cl current inhibition by anion transport blockers. A: examples of single channel current traces recorded using cell-attached configuration from a holding potential of 55 mV in control and after application of 100 µM niflumic acid. Pipette solution contained 2 mM Ca2+. Test potential used was +60 mV. Niflumic acid completely abolished the single channel current activities. B: examples of single channel current records obtained using quartz electrodes in symmetrical Cl recording solution. Patches were recorded during cell-attached and excised inside-out configurations into external solution with pCa of 3. Pipette solution contained no Ca2+, and outward currents can be seen only after excision of the patch into the external solution. These currents can be blocked by DIDS (100 µM) applied directly to the bath solution. Similar data were obtained in a total of 5 patches. Open probability in control and after application of DIDS is shown (bottom).

Functional significance of Ca²+-activated Cl current. To establish the functional significance of the niflumic acid-sensitive current, action potentials were recorded at room temperature using the perforated-patch technique at different stimulation frequencies in the control and after application of niflumic acid (Fig. 8). At a low stimulation frequency, application of niflumic acid resulted in only slight shortening of the terminal phase of the action potential with no observable effects on the action potential profiles at faster stimulation frequencies (Fig. 8A). The slight shortening of the terminal phase of the action potential can be attributed to blockade of the inward component of I_Cl,Ca. The calculated E_Cl in our recording condition was approximately 40 mV. In contrast, in the presence of 5 mM 4-AP to block the Ca²⁺-insensitive I_to, niflumic acid led to a significant prolongation of the early repolarization phase of the action potential. The effect became more pronounced at faster stimulation frequency (Fig. 8B). In addition, at low stimulation frequency, the modest shortening of the terminal phase of the action potential could still be observed. Summary data are shown in Fig. 8C, which depicts changes in action potential duration at 50% and 90% repolarization (APD₅₀ and APD₉₀, respectively). Figure 8D shows the effects of 4-AP and 4-AP plus low-[Cl]_o solution on the action potential profile, which further confirms the contribution of Cl current on the repolarization of the action potential.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Functional contribution of ICl,Ca at different stimulation frequencies. A: action potentials recorded from a control cell using a perforated-patch variant at room temperature at stimulation frequencies of 0.2, 0.5, 5, and 10 Hz in control and after application of 100 µM niflumic acid. Membrane potential at 0 mV is indicated (solid line, left). B: action potentials recorded from a second cell in control and after administration of 5 mM 4-AP and 5 mM 4-AP plus 100 µM niflumic acid. C: summary data showing the action potential duration at 50% and 90% repolarization (APD50 and APD90, respectively) in control and after administration of 4-AP and 4-AP plus niflumic acid at different stimulation frequencies. Resting membrane potential was unchanged after drug applications. Data represent means ± SE, n = 6 cells from each group; * P < 0.05 compared with 4-AP alone; # P < 0.05 compared with control. D: action potential recorded in a control cell, after 5 mM 4-AP administration, and after 4-AP administration plus low [Cl]o (6 mM) at a stimulation frequency of 5 Hz.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anion channels in the heart have been shown to mediate a variety of functions and thus play a potentially important role in cardiac physiology and pathophysiology (20, 44). Activation of the anion channels can significantly alter resting membrane potential and the duration of the action potential. These proteins may represent novel targets for the development of new antiarrhythmic drugs. Furthermore, new exciting data have shown linkages of several human genetic diseases to specific anion channel defects (1, 30). However, our present understanding of the physiological significance of the various anion channels in the heart remains incomplete.

Here we reported the presence of an I_Cl,Ca in mouse ventricular myocytes. The activation of the channel is critically dependent on Ca²⁺ entry via voltage-gated Ca²⁺ channels and release from intracellular Ca²⁺ stores. Similar to that described in canine cardiac myocytes, the channel showed a small single channel conductance of ~1 pS (6). On the other hand, we have demonstrated several important differences. In contrast to the previously described cardiac I_Cl,Ca, an important finding from our study was that the I_Cl,Ca in mouse ventricular myocytes also exhibited voltage dependence. In addition, we have demonstrated that the repolarization process in mouse ventricular myocytes is significantly different from that of dogs or larger animals. In mouse ventricular myocytes, the repolarization of the action potentials is dominated by the anomalously large 4-AP-sensitive I_to. Indeed, the presence of I_Cl,Ca could only be well appreciated when I_to is blocked. These findings may have significant implications when transgenic mouse models are used to study the repolarization process in the heart; e.g., mouse models of long QT syndrome (5, 29).

Single channel studies. The existence of I_Cl,Ca in mouse ventricular myocytes is substantiated by the findings of the sensitivity of the channels to Cl concentrations and the dependence of the channels to Ca²⁺ entry via voltage-gated Ca²⁺ channels and Ca²⁺-induced Ca²⁺ release (CICR). In addition, the channel could be blocked by anion transport blockers. However, one previous study (50) has indicated that anion transport blockers and caffeine are nonspecific and could affect voltage-gated K⁺ channels Kv4.2 and Kv4.3, which underlie (at least in part) the Ca²⁺-insensitive component of the I_to. Our single channel studies directly ruled out this possibility. We were able to document a direct blockade of the single channel activities by niflumic acid and DIDS. Furthermore, similar to a previous report in canine ventricular myocytes (6), the I_Cl,Ca identified exhibits a very small single channel conductance (1.0-1.3 pS). Indeed, the single channel conductance is comparable to the low-conductance Ca²⁺-activated Cl channels previously reported in Xenopus oocytes (46) and smooth muscle cells (49). Despite the low single channel conductance, cardiac I_Cl,Ca was reported to have a high membrane density (~3 channels/µm²) in canine cardiac myocytes (6). Similarly, in our present study, we estimated the channel density to be almost the same value (2.5 channels/µm²) in mouse ventricular myocytes assuming a specific membrane capacitance of 1 µF/cm² (18), an estimated single channel current amplitude of 0.12 pA (in cell-attached configuration), a single channel open probability (P_o) of 0.05, and a macroscopic current density of 1.5 pA/pF at +60 mV.

Dependency of I_Cl,Ca on Ca²+ entry via voltage-gated Ca²⁺ channels and CICR. Similar to previous studies (42, 55), the activation of I_Cl,Ca in mouse ventricular myocytes is critically dependent on Ca²⁺ entry via voltage-gated Ca²⁺ channels and CICR. The presence of a subsarcolemmal microdomain of Ca²⁺ has previously been documented (32, 45). The Ca²⁺ concentration within the subsarcolemmal microdomain is estimated to be significantly higher than that of the bulk cytoplasmic Ca²⁺ concentration during CICR. One recent study (47) has shown that I_Cl,Ca and the Na⁺/Ca²⁺ exchanger current have different time courses during SR Ca²⁺ release in ferret ventricular myocytes. It was suggested that the Ca²⁺-activated Cl channel might be tightly coupled to the sarcolemmal Ca²⁺ channel as well as the ryanodine-release channel (47). Thus one would predict that the activity of the channels could be closely regulated by multiple factors known to modulate sarcolemmal Ca²⁺ channels and the CICR; e.g., -adrenergic signaling.

Species and tissue distribution. I_Cl,Ca has been studied mostly in rabbit atrial, ventricular (56, 57), and Purkinje cells (42) and canine ventricular myocytes (6, 48, 54, 55). It has also been detected in sheep cardiac Purkinje fibers (24) and cultured chick cardiac cells (33, 34). However, it appears to be absent in guinea pig ventricular myocytes (43). In human atrial and ventricular myocytes, the I_to is considered to be one of the major repolarizing currents (36, 53). However, the existence of a Ca²⁺-sensitive component of the I_to in human heart remains controversial (20, 44). Early studies have documented the presence of the current in atrial tissue (12); however, a more recent study (31) found that the 4-AP-resistant component of the I_to was Ca²⁺ insensitive and suggested that the current may represent the voltage-dependent relief of 4-AP block of the transient outward K⁺ current.

Molecular correlates of I_Cl,Ca. At least four primary types of sarcolemmal Cl channels have been described in the hearts. The small-conductance Ca²⁺-activated Cl channel is by far the most ubiquitous across different cell types. However, the exact molecular identification of this channel in the heart remains unknown. A novel family of Ca²⁺-activated Cl channels (CLCA) has recently been discovered with multiple members being expressed in different tissue and species (39). So far, two bovine, three mouse, and four human CLCA family members have been cloned. Each CLCA exhibits a distinct and often overlapping tissue-expression pattern. The clones from bovine trachea (bCLCA1; Ref. 9) have a relatively large unitary conductance and are insensitive to niflumic acid. More recently, proteins with homology to bCLCA1 have been cloned from a mouse lung cDNA library (mCLCA1; Ref. 13) and from a human genomic library (hCLCA1 and hCLCA2; Refs. 14 and 15). mCLCA1 can be found in a variety of tissues including the heart. The expression of hCLCA1 appears to be specific only to intestinal cells, whereas hCLCA2 is found in the lung and trachea. Additional studies are required to further confirm the molecular identity of the Ca²⁺-activated Cl channels in cardiac myocytes.

Physiological and pathological significance. Several functional roles have been suggested for I_Cl,Ca: the current may exert a negative-feedback mechanism on the voltage-gated Ca²⁺ channel by limiting the action potential duration during the plateau phase (20). In addition, I_Cl,Ca is tightly coupled to the process of CICR. Therefore, the current can be expected to be modulated by an increase in -adrenergic stimulation or a decrease in muscarinic receptor stimulation as a direct result of changes in intracellular Ca²⁺. Under certain conditions, I_Cl,Ca can be activated via CICR triggered by the Na⁺/Ca²⁺ exchanger operating in the reverse mode (27). In canine cardiac myocytes, I_Cl,Ca was found to be important in early repolarization (phase 1) especially during fast stimulation frequency (59). In this study, significantly different findings were observed in mouse ventricular myocytes. The early repolarization in mouse cardiac myocytes is dominated by the large 4-AP-sensitive I_to; therefore, blockade of the I_Cl,Ca has no observable effects on the early repolarization of the action potential profile. Only a slight shortening of the terminal phase of the action potential is observed at low stimulation frequency. In contrast, in the absence of the 4-AP-sensitive I_to, blockade of I_Cl,Ca leads to a significant prolongation of the action potential. A small but significant shortening of the terminal phase of the action potential could still be appreciated at low stimulation frequency. Indeed, our present study suggests some of the difficulties one may face in using the murine models. For example, one recent study shows no significant cardiac abnormalities in a mouse model with targeted ablation of the Kcnq1 gene (29), whereas a second study shows a prolongation of the QT interval in the same model (5).